This relates generally to image sensors, and more specifically, to methods and circuitry for operating image sensor pixels with dual-gain readout for producing high dynamic range (HDR) images.
In conventional imaging systems, image artifacts may be caused by moving objects, moving or shaking camera, flickering lighting, and objects with changing illumination in an image frame. Such artifacts may include, for example, missing parts of an object, edge color artifacts, and object distortion. Examples of objects with changing illumination include light-emitting diode (LED) traffic signs (which can flicker several hundred times per second) and LED brake lights or headlights of modern cars.
While electronic rolling shutter and global shutter modes produce images with different artifacts, the root cause for such artifacts is common for both modes of operation. Typically, image sensors acquire light asynchronously relative to the scenery being captured. This means that portions of an image frame may not be exposed for part of the frame duration. This is especially true for bright scenery when integration times are much shorter than the frame time used. Zones in an image frame that are not fully exposed to dynamic scenery may result in object distortion, ghosting effects, and color artifacts when the scenery includes moving or fast-changing objects. Similar effects may be observed when the camera is moving or shaking during image capture operations.
Conventional imaging systems also may have images with artifacts associated with low dynamic range. Scenes with bright and dark portions may produce artifacts in conventional image sensors, as portions of the image may be over exposed or under exposed.
Dual gain pixels are commonly used to improve the dynamic range of an image sensor. They can be used either in a fixed high or fixed low gain readout mode or in a dual readout mode where both gain modes are read out. In the dual readout mode, charge is either stored entirely on the photodiode or is allowed to overflow to a floating diffusion node during integration. The combination of dual gain readout with overflow during integration allows for the largest dynamic range increase.
Dual gain pixels traditionally read out captured high-gain and low-gain image data in respective high-gain and low-gain configurations. Switching between the high-gain configuration and the low-gain configuration results in electrical crosstalk. This crosstalk causes an undesirable large electrical offset between signals read in the high-gain configuration and signals read in the low-gain configuration. This electrical offset can cause pixel output signals to have a magnitude that is outside of the operating range of analog readout circuitry in the imaging system.
Dual gain pixels traditionally read out captured image data using a method that requires either four pixel read operations and analog to digital conversions (ADCs) to operate without a frame buffer, or three pixel reads and three ADCs to operate with a frame buffer. In the latter case, the frame buffer is required to provide a reference image for offset correction between signals. Performing additional reads and ADC conversions requires additional power, whereas the introduction of a frame buffer requires additional hardware. Such increased power consumption and addition of hardware is generally undesirable.
It would therefore be desirable to be able to provide high dynamic range (HDR) image sensors that do not have a large electrical offset between pixel output signals, that require fewer reads and ADC conversions than traditional image sensors, and that do not require a frame buffer for pixel readout.